Most Regulatory Records Are Paperwork. Some Are a Client Who Lost Money.

Every investment advisory firm in the United States must give clients a document explaining what it charges and where its interests might conflict with theirs. It is called Form ADV Part 2. It is free, it is public, and almost nobody reads it.

Most firms have a clean record. On the records that do exist, a late filing, a customer complaint and a case where clients lost money all look the same from the outside. Each one is “an event.” That is what makes a regulatory record hard to read, and it is why the distinction above matters.

The document is not associated with whether a firm has had problems.
It is associated with what kind.

What we found

About one firm in eleven — 1,772 of them — has something on its public record. Some of those entries describe an affiliated company or an individual adviser rather than the firm itself. The filings only mark severity on a firm’s own conduct, so the comparison below uses the 954 firms whose own conduct is on the record. We sorted those by how clearly they had written their fee and conflict disclosures, then looked at what their events actually were.

The filings themselves mark some matters differently from the rest. A client was harmed — someone lost money. The firm failed to supervise — it did not catch what one of its own people was doing. Or fraud was involved. Everything else is closer to administrative: a form filed late, a disclosure updated behind schedule.

Those three marks are what we counted.

The groups in the table are Fidelon scores. Every firm we cover is scored 0–100 on how clearly it discloses what it charges and where its interests diverge from its clients’ — part of the Fidelon TransparencyScore, and visible on a firm’s profile under Filing Quality. A firm scoring 80 or above explains its fees in language a client can follow. A firm below 60 does not. The bands below are those scores, sorted.

Read the first column down: 5.6%, 10.7%, 12.9%, 27.5%. It rises at every step, and the step widens as the writing gets harder to read. Firms in the hardest-to-read group were 4.9times as likely to have an event where clients lost money as firms in the clearest.

Or read it the other way: 94.4% of the clearest-scoring firms with a record had no client losses on it. Writing a brochure a client can understand is not rare, not expensive, and not beyond any firm in this study. Thousands of them did it.

Stated across all 20,092 firms

The table above describes firms that already have a record of their own conduct. Most firms do not. Applied to the full population, the same finding reads:

About one firm in 582 with the clearest disclosures has a customer-harm matter on its record. About one in 95 with the hardest to read. The ratio is roughly the same, 6.1×. The absolute numbers are small, and we report them so that no reader mistakes the figures above for a statement about firms in general. They are a statement about records, not about firms.

Does firm size explain it?

Partly, and we will not paper over where it does.

The severity pattern holds inside the two larger size classes and is significant in both:

At firms with more than fifty advisors, 55.9% of the least clear had a matter where clients lost money, against 16.7% of the clearest. That top cell rests on 12 firms, so read it as direction rather than a precise rate — the gradient below it does not depend on that one number.

Below eleven advisors, it falls apart. The groups get small, the rates bounce around, and claiming a pattern there would be inventing one. For a solo advisor or a two-person shop, this study does not tell you much. We would rather say so than let the headline imply otherwise.

Does an unclear brochure mean a firm is more likely to be in trouble at all?

No — and we tested this first, expecting the opposite.

Across all 20,092 firms, the share with anything at all on their record does not rise as disclosures get harder to read. It runs 6.1%, 11.2%, 10.9%, and 6.2% from clearest to least clear. That is not a gradient. The firms with the hardest-to-read brochures are no more likely to have a regulatory record than the firms with the clearest.

What we tested that did not hold up

One thing, and it is the version of this finding we most wanted to be true. We report it because a study that only publishes what worked is advertising.

• Whether the document points forwardThe useful version of this finding would be predictive — clear writing today, fewer problems tomorrow. We tested it: we took firms with no events before a cutoff date and asked which ones picked up a first problem afterward. The pattern broke. Firms in the middle group had the highest rate of new problems, and the least clear group came in below them. We tried four different cutoff years and it broke the same way each time. We cannot say this document predicts anything, and we are not going to say it.

What comes next

This is the first of these. The finding above is one relationship in a dataset that holds far more — every registered firm’s filings, disclosures, and public record. We expect to find patterns that are sharper than this one, and some that contradict what we currently believe. We will publish both.

What this study cannot tell you

• It is an association, not a cause. Clear writing does not prevent regulatory problems. The most plausible reading is that both come from the same source: whether a firm takes its compliance obligations seriously enough to fund them properly. A firm with a real compliance function tends to produce a readable brochure and tends to keep its problems administrative. The brochure is the part of that you can see from outside, without asking anyone's permission.
• It describes firms in aggregate. It supports no conclusion about any individual firm, and Fidelon does not apply it to one. A firm's score reflects what that firm disclosed. It is not a forecast of that firm's conduct.
• It is a larger-firm result. The pattern is statistically reliable at firms with eleven or more advisors. Below that — solo practices and small partnerships, which are most firms — the rates bounce around and we cannot claim a pattern. The headline figures pool all sizes together, so they lean on the larger firms.
• A regulatory record only shows what someone reported. It captures matters that were surfaced and filed. It cannot show conduct that was never caught, and the chance of being caught may differ between firms in ways we have not measured. About 27% of the events on file also carry no date.
• Older matters drop out. We include only events still within their reporting lifecycle. A firm whose problems have aged off the record appears here as clean.
• Attribution involves judgment. Some events in the underlying record are not yet assigned to a responsible party, and some are placeholder entries. Both are excluded here.
• One bias runs in our favor, and we will name it. After a regulatory action, a firm's lawyers typically rewrite its disclosure language. That would push a troubled firm's clarity score up and flatten exactly the pattern we found. If that effect is present, the real relationship is stronger than what is reported here, not weaker.

How to read your own advisor’s brochure

Nothing above tells you anything about your own advisor. It describes 20,092 firms in aggregate, and a single firm is not a population.

What it does suggest is that the document is worth five minutes.

1. 1Ask your advisor for their Form ADV Part 2, or search the firm’s name at adviserinfo.sec.gov. It is free and public.
2. 2Find the section on fees — usually Item 5.
3. 3Read it once, at normal speed, the way you would read anything else.
4. 4Ask yourself one question: can I tell what I would actually pay? Not the percentage — the dollars, including what the funds inside your account charge, what you pay for custody and trading, and what the firm receives from anyone other than you.

If you can tell, that is a firm that decided you should be able to. If you cannot, after an honest read, that is worth noticing — and worth asking about.

It may mean nothing at all. Most firms with hard-to-read brochures have clean records. But it is the same thing we measured.

Questions worth asking

Each has a short, specific answer any firm should be able to give without preparation.

1. 1

   In dollars, what did I pay you last year — including anything charged inside the funds I hold?

   A firm that knows its own fee structure can answer this. The answer should be a number.

2. 2

   Does anyone other than me pay you anything in connection with my account?

   This covers revenue sharing, commissions, and payments from fund companies. The answer should be either "no" or a specific list.

3. 3

   Are you registered as a broker-dealer as well as an investment adviser?

   If yes, ask which standard applies to you and when. Advisers owe a fiduciary duty — they must put your interests first. Brokers owe suitability, a lower bar. Many people are both, at different moments, in the same meeting.

4. 4

   Which section of your Form ADV Part 2 answers question 1, and can you point me to it?

   This is the useful one. It asks the firm to show you its own disclosure.

5. 5

   Has the firm reported any regulatory or customer complaint matter in the last ten years?

   Whatever the answer, you can verify it yourself on the SEC's public database.

Methodology

Peer set

20,092 SEC- and state-registered investment advisory firms — every firm whose Form ADV Part 2 fee and conflict disclosures Fidelon has scored. No firm with a clarity score was excluded. Nothing is estimated or imputed. Firms registered as both investment advisers and broker-dealers file the same brochure and are included. They are unevenly distributed across the bands and carry a higher client-harm rate, so they contribute more to the lowest band than to the highest.

Clarity scoring

Fidelon scores the fee and conflict-of-interest sections of every firm's Form ADV Part 2 on a 0–100 scale, for whether a reader can determine what they would pay and where the firm's interests diverge from theirs. It combines published readability measures with pattern analysis that identifies language constructed to obscure rather than explain. The measure is part of the Fidelon TransparencyScore and surfaces on a firm's profile under Filing Quality. It does not read disciplinary history, Item 9, Item 11, or any disclosure record. It does not read firm size, assets under management, or registration type. This is what makes the comparison meaningful: the score cannot contain the answer.

Event data

Disclosure Reporting Pages (DRP — the standard form a firm files to report a regulatory action, customer dispute, or similar matter) as filed, deduplicated at the firm level. Events still within their reporting lifecycle only. Events whose responsible party is undetermined, and placeholder records, are excluded. An event counts if it involved the firm, an affiliated company, or an advisor working there.

Severity markers

Whether a firm has at least one event flagged in the filings for customer harm, supervision failure, or fraud. These flags come from the filings themselves, not from our interpretation of them. They are recorded only on events attributed to a firm's own conduct, so the severity tables use a matched denominator: the 954 firms with at least one such event, not the 1,772 with anything at all on record. Reported against the wider denominator the same gradient runs 2.8% to 17.1% (6.1×) — same direction, lower rates, because firms that cannot appear in the numerator sit in the denominator. These flags are also applied at different rates by different regulatory data sources, so this study is confined to firms whose records come through Form ADV, which is why it covers investment advisers and not broker-dealers.

Score bands

The four groups used throughout this note are bands of the Fidelon disclosure clarity score described above — 80 and over, 70–79, 60–69, and below 60 on the 0–100 scale. They are the same scores shown on a firm's profile, not a grouping created for this study. The cutoffs were fixed before the analysis was run and are not adjusted between revisions.

Statistics

95% confidence intervals on all proportions (Wilson). Monotone trends tested with the Cochran-Armitage test, two-sided: customer harm p = 1.4 × 10⁻⁷, supervision failure p = 2.3 × 10⁻¹⁰, fraud p = 8.1 × 10⁻⁶. P-values are unadjusted for multiple comparisons. Groups with fewer than 20 firms are suppressed rather than reported, and cells resting on small counts are labelled where they appear.

Data vintage

Brochures: half filed after 11 September 2025; 90% filed between March 2025 and July 2026. Events: through 28 July 2026.

Reproducibility

The full analysis specification — peer-set filter, both event definitions, deduplication rule, group construction, and test design — is published. The component weights inside the Fidelon TransparencyScore are not; the principle each component measures is. We consider a score whose reasoning you can audit, but whose arithmetic you cannot game, the correct trade.

Data sources

Form ADV Parts 1, 2, and 3 (Form CRS) as filed with the SEC and made available through Investment Adviser Public Disclosure (IAPD). Disclosure Reporting Pages as filed. All underlying data is public and can be retrieved by any reader.

The firms that wrote clearly

6,399 firms in this study scored 80 or above — they wrote fee and conflict disclosures a client can actually read. Among those that had a record, 94.4% had no client losses on it.

They did that before we existed and without being asked. Most of them will never know they were measured.

The industry’s opacity is architectural — built out of hundred-page filings, layered compensation, and disclosure written by lawyers for regulators rather than by firms for clients. The firms above are the evidence that the architecture is a choice rather than a constraint.